https://github.com/sourdoughcat/scene-completion
A Python Implementation of "Scene Completion Using Millions of Photographs"
https://github.com/sourdoughcat/scene-completion
Last synced: 6 months ago
JSON representation
A Python Implementation of "Scene Completion Using Millions of Photographs"
- Host: GitHub
- URL: https://github.com/sourdoughcat/scene-completion
- Owner: SourdoughCat
- Created: 2015-11-16T10:12:06.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2015-12-13T08:21:55.000Z (over 9 years ago)
- Last Synced: 2024-12-09T00:10:15.707Z (6 months ago)
- Language: Jupyter Notebook
- Size: 62.1 MB
- Stars: 20
- Watchers: 1
- Forks: 6
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
Awesome Lists containing this project
README
Readme
======
This is a naive Python implementation of Local Context Matching as shown in [Scene Completion Using Millions of Photographs](http://graphics.cs.cmu.edu/projects/scene-completion/) which was completed as part of Georgia Tech's Computational Photography Course.
This repository contains the Jupyter notebook I worked off as well as the functions I used to generate the local context matching. More information of the approach can be found [here](https://docs.google.com/presentation/d/1ObIpms39d0bY6UPnAt8woTY66nbgHukCCf_94OQjhtQ/pub?start=false&loop=false&delayms=3000).
Additional Outputs are located in the following albums:
*
*
*
Other Notes and Work
--------------------
The Flickr Download code is based on the code ["Download Images from Flickr with Python"](http://halotis.com/download-images-from-flickr-with-python/)
Whilst my (failed) attempt at the gist descriptor is based on the hints provided on [quora](https://www.quora.com/Computer-Vision/What-is-a-GIST-descriptor). The GIST descriptor is used to perform similar image matching within the [Scene Completion paper](graphics.cs.cmu.edu/projects/scene-completion/).
The sample images are based on the [Scene Completion work assignment which was implemented in Matlab](http://cs.brown.edu/courses/cs129/results/final/zyp/)
Notes
-----
The Gist descriptor for the three channels looks at all the colours rather than the grayscale images only.